834 M. Arago on the Colour of the Ocean, 



equals 37°, they are polarized. A piece of tourmaline will then 

 completely eliminate them, and thus the blue of the sea will be 

 seen apart without any extraneous mixture. 



With the intention of being freed, as much as possible, from 

 the influence of reflection when studying the colours of the 

 ocean, some able navigators have recommended that we should 

 always examine it through the trunk of the ship's rudder. In 

 this method the water exhibits, in some points of view, beautiful 

 violet tints ; but with a little attention we may be convinced 

 that these hues are not at all real ; that they are only the effect 

 of contrast ; that they proceed from atmospheric light feebly 

 reflected in an almost perpendicular direction, and coloured 

 by their approximation to the transmitted green colours which 

 almost invariably surround the rudder. 



If there be some who wish to admit, and develope, this at- 

 tempt at an explanation of the colours of the sea, which we 

 have just given, or if there be others who wish to refute it, and 

 substitute a better in its place, it will still be necessary to be- 

 gin by investigating the colour of water when seen by trans- 

 mission zaith the aid of diffused light. Those who will recall to 

 recollection the pre-eminently green hue of the cut edge of a 

 crystal glass, even when this glass is only illuminated in front 

 and perpendicularly, will perceive all the importance of this re- 

 mark. The following appears a very simple means of reaching 

 a satisfactory conclusion. 



I shall suppose that the observer is supplied with one of 

 those large hollow ice prisms, which philosophers are in the habit 

 of using when they wish to study the refraction of liquids. To 

 make our apprehensions the more precise, we shall make the re- 

 fracting angle equal to 45°, we shall then suppose that the 

 prism is plunged 'partially into water, so that the edge of its 

 refracting angle shall be downwards and horizontal, and that 

 one of the faces of this angle, viz. that which is most exposed 

 (est tournee vers le large), shall be vertical, whence it will 

 result as a necessary consequence that the other face will be 

 inclined to the horizon at an angle of 45°. 



When the objects are thus arranged, the light which moves 

 horizontally in the water at a fraction of an inch below the sur- 



